The major goal of this proposal is to establish and elucidate the pathogenesis of experimentally induced simian immunodeficiency virus infection in rhesus monkeys as a model for AIDS in humans. We will inoculate rhesus monkeys subcutaneously with virus and follow the course of infection, paying particular attention to mechanisms of spread of virus among tissues, from lymphatic tissues and bone marrow to the central nervous system and in the interaction of virus with macrophages and lymphocytes. Tissues (and cells) that become latently or productively infected will be identified using infectivity assays of plasma and tissue extracts and infectious center assays of mononuclear cells from peripheral blood. Synthesized viral polypeptides will be identified by immunocytochemistry in tissue sections. Viral RNA transcripts will be identified by in situ hybridization using isotopically labeled viral DNA probes. Particular cells that are producing virus will be identified by combined immunocytochemistry for the demonstration of cellular phenotypes and in situ hybridization for demonstration of viral RNA transcripts. Latently infected cells expressing small numbers of copies of proviral DNA will be identified by viral gene amplification procedures. Aberrant expression of host genes such as MHC antigens, interferons or tumor necrosis factor will be identified by immunocytochemistry and/or by hybridization using isotopically labeled probes. Using virus and plasma collected during the persistent infection in monkeys, we will perform neutralization tests to determine the frequency and extent of antigenic variation of the virus in different animals. We will dissect the mechanism of neurotropism of the virus by identification of cells in the CNS and PNS that support replication of the virus. In order to address a hypothesis that neurotropic viruses arise by mutation and selection of virus, we will passage cloned non-neurotropic virus sequentially in cultures of monkeys macrophages and in brain by sequential intracerebral inoculation of monkeys and select for neurovirulent virus. Neurotropic/neuroinvasive properties of these viruses will be characterized biologically by determination of the routes of invasion of the CNS and identification of the cells that support replication. Pathogenesis of the CNS lesions will be assessed by comparing viral gene expression in particular cells types with aberrant expression of host genes in these or neighboring cells.